English
▼
Supercharge your LLM with Sudoku100's AI-native toolchain — MCP Server and Skill Package. Let any AI model generate, embed, and deliver Sudoku puzzles through natural language or structured function calling.
MCP Server for Claude / Cursor Skill Package for OpenAI / LangChain No API key, 100% free
Get Started
MCP Server
One command to connect Claude Desktop, Cursor, or any MCP-compatible client to Sudoku100 puzzles.
npx -y sudoku-api-mcp{
"mcpServers": {
"sudoku": {
"command": "npx",
"args": ["-y", "sudoku-api-mcp"]
}
}
}Skill Package
Drop into any Node.js project. OpenAI Tool Calls, Claude tools, LangChain — universal LLM function calling.
npm install sudoku-api-skillconst skill = require('sudoku-api-skill');
const tools = skill.getToolDefinition();
// → [{ type: "function", function: { ... } }]
const result = await skill.invoke({
action: "generate_sudoku",
difficulty: "medium",
width: 400,
format: "png"
});
// → { success: true, imageUrl: "...", puzzleId: 12345 }MCP Server
{
"mcpServers": {
"sudoku": {
"command": "npx",
"args": ["-y", "sudoku-api-mcp"]
}
}
}Model Context Protocol (MCP) server that brings Sudoku generation directly into any MCP-compatible AI client. Connect once, generate puzzles forever.
- generate_sudoku Generate a Sudoku puzzle image by difficulty. 6 levels from beginner to extreme.
- get_sudoku_by_id Retrieve a specific Sudoku puzzle by its numeric ID. Same ID = same puzzle.
- list_difficulties List all available difficulty levels with hint count ranges.
Skill Package
An OpenAI-compatible tool definition and executor. Add sudoku_api as a tool to any LLM that supports function/tool calling — OpenAI, Claude API, Gemini, LangChain, and more.
Compatible with: OpenAI · Claude API · Gemini · LangChain · LlamaIndexconst skill = require('sudoku-api-skill');
// Get OpenAI-compatible tool definition
const tools = skill.getToolDefinition();
console.log(JSON.stringify(tools, null, 2));
// Output:
[{
"type": "function",
"function": {
"name": "generate_sudoku",
"description": "Generate a Sudoku puzzle",
"parameters": {
"type": "object",
"properties": {
"difficulty": { "type": "string" },
"width": { "type": "number" },
"format": { "type": "string" }
},
"required": ["difficulty"]
}
}
}]const result = await skill.invoke({
action: "generate_sudoku",
difficulty: "medium",
width: 400,
format: "png"
});
// Result:
// {
// success: true,
// imageUrl: "https://sudoku100.com/...",
// puzzleId: 12345,
// difficulty: "medium"
// }3 Ways to Use Sudoku100 with AI
Choose the integration method that fits your workflow best.
1
Direct URL
No installation needed. Just paste a Sudoku100 URL into any AI chat or tool. The simplest way to get started.
https://www.sudoku100.com/sudoku-img 2
MCP Server
Best for Claude Desktop and Cursor users. Install once, then simply ask AI to generate a Sudoku in natural language.
npx -y sudoku-api-mcp 3
Skill Package
For developers building AI apps. Add sudoku_api as a function tool in your LLM pipeline with full parameter control.
npm install sudoku-api-skill Use Cases
AI-powered Sudoku integration opens up a new world of possibilities.
AI-Enhanced Education
Let LLMs generate personalized Sudoku worksheets for students at the right difficulty level. Great for math enrichment and logic training.
Smart Content Creation
Bloggers and newsletter authors can have AI automatically insert fresh Sudoku puzzles into their articles and emails.
In-App Game Integration
Add Sudoku to your app with minimal code. AI handles puzzle generation while you focus on app experience.
Data Analysis & Testing
Researchers can use LLMs to batch-generate Sudoku datasets with controlled difficulty distributions for studies.
Chatbot Enrichment
Make your chatbot more engaging by letting it offer Sudoku challenges as a fun break during conversations.
Multi-Agent Workflows
Combine Sudoku generation with other MCP tools to build complex multi-step AI automation pipelines.
Why Sudoku100 for AI?
Purpose-built for the AI era — our MCP Server and Skill Package make Sudoku generation a first-class LLM capability.
Zero Authentication
No API keys, no sign-up, no tokens. Both MCP and Skill work out of the box. Perfect for AI agents that need frictionless access.
MCP-Native Design
Full Model Context Protocol support. Connect Claude Desktop or any MCP client with a single JSON config block.
OpenAI Tool Call Compatible
Skill package implements standard OpenAI function calling schema. Use with any LLM that supports tools/functions.
Guaranteed Unique Solution
Every puzzle is validated through rigorous solving algorithms. No ambiguous or broken puzzles — AI can always deliver quality.
6 Difficulty Levels
Beginner to extreme. 6 refined tiers let AI tailor puzzle difficulty to the user's exact needs.
Multiple Output Formats
PNG, WebP, SVG, JPG with customizable width (100-1000px). AI can pick the right format for any context.
Deterministic Puzzle IDs
get_sudoku_by_id always returns the same puzzle. Perfect for reproducible AI outputs and shareable content.
Open Source & Free Forever
MIT licensed. No usage limits, no hidden fees. Built by the Sudoku100 community for the AI community.
FAQ
What's the difference between MCP and Skill?
MCP (Model Context Protocol) is a protocol for AI clients like Claude Desktop to discover and call tools automatically. The Skill package is a JavaScript library you import into your own code to add Sudoku as a function tool to any LLM pipeline. Use MCP if you're an end-user of AI clients; use Skill if you're a developer building AI apps.
Do I need both MCP and Skill?
No, you only need one. MCP is for Claude Desktop/Cursor users who want natural language access. Skill is for developers adding Sudoku to their own AI applications.
Is the MCP server free?
Yes! The MCP server is completely free, just like our public API. No registration or API key required.
Which AI platforms are supported?
MCP Server: any MCP-compatible client (Claude Desktop, Cursor, Continue, etc.). Skill Package: OpenAI, Claude API, Gemini, LangChain, LlamaIndex, and any LLM with function/tool calling support.
Can I customize puzzle size and format?
Absolutely! Both MCP and Skill support width (100-1000px) and format (png, webp, svg, jpg) parameters. Just specify them when calling the tool.
Are there rate limits?
No. There are no rate limits on MCP or Skill usage. Generate as many puzzles as your AI workflow needs.
Where can I find the source code?
Both MCP server and Skill package are open source on GitHub at github.com/sudoku100com-create/Sudoku-generate-API. Stars and contributions welcome!
How do I update the MCP server?
Just run npx -y sudoku-api-mcp again — the -y flag ensures you always get the latest version from npm.
